The Web At A New Crossroads


Guest post by Chris Messina and Jyri Engeström (thanks to Brynn Evans for editing and Brad Fitzpatrick for comments on the draft)

Around 2003, things began to change.

Technology was then the black sheep, having left overnight millionaires destitute and without change to afford their $4 lattes. Even the posers had left San Francisco and gone back to suburbia to be office managers at Walmart.

It was a sad time for everyone — that is, except the die-hards and the hackers. The web for them had never been about making money, but about reshaping culture and toppling the old order. 2003, therefore, was the perfect time for a resurgence: the people who kept pushing on in the Valley and elsewhere were a concentrated motley crew of innovators and builders. They cared about technology for technology's sake and about developing and advancing web culture.

What they didn’t realize, however, was that the services and technologies that they were destined to build would need to be cobbled and sewn together using a system that would fight them every step of the way — not out of spite — but because of its architecture. By definition the network available was decidedly anti-human: in 2003, there was only the document-centric web.

The document-centric web

We’ll spare you the history lesson of the origin story of the internet, but suffice it to say, the web we have today is because a bunch of scientists, academics, and government folks needed a way to share static documents — not set up identities or have a dynamic conversation in public. The net was decidedly antisocial and anti-serendipity, from the beginning.

Keep that in mind when you consider what happened around 2003: masses of people started blogging, publicly. Services like Blogger and TypePad surged; LiveJournal and WordPress started to grow stubble and Drupal emerged from a college dorm. In the absence of innovation since the bubble burst, people started to realize that the web could be a place for personal expression and public conversation — and blogging became the “it” thing to do.

The problem was that tools were built around the document model of publishing. Many people maintained collections of blogs that they kept handy as bookmarks — and visited regularly, sometimes several times a day (depending on the prolificness of a given blogger). The more savvy audiences discovered desktop feed readers that fetched new content automatically. But conversation was fragmented and inconvenient: to comment, you had to visit the publisher’s blog and create a single-purpose account there; to post an original response, you had to have your own blog and know how to send a trackback to the post you were responding to.

The pace was slow and cumbersome, but most early bloggers didn’t mind. Their new medium was exciting, expansive, and controversial. And for the time, it fit the write-print/publish model many people had become familiar with thanks to Microsoft Word and other text editors — and which was in turn rewarded by Google’s link-based approach to search.

But two things were lacking in the first generation of Web 2.0 tools: personhood and aggregated conversation streams. The document-web hadn’t made room for people-friendly affordances like “faces,” and didn’t conform to our restless animal brain, which is well suited to working with a flow of short snippets of information.

Proprietary, real-time platforms

Enter: the real-time web. If 2003–2006 could be defined as the emergence of social media on infrastructure still dominated by the document-web, 2007 through the present will be defined as the transition to the "real-time" web, even if through a proprietary side-road.

We’ve had chat, SMS, and other forms of asynchronous (near) real-time data streams for some time. But, just as blogging did to email, every new generation is about pushing down the walls that cage one-to-one and one-to-few interactions, turning the same private publishing tools into many-to-many-to-many-more public publishing platforms. Emphasis on the noun: from tools to platforms.

The catch? This real-time web is not mature yet, since the platforms that sequester all of our activities today are proprietary ones like Facebook and Twitter. These are convenient, to be sure, but of limited utility to users with cross-site ambitions, who require interoperability.

While “brand-mediated" profiles and relationships may not seem completely odious on the surface, there are four major drawbacks to keep in mind:

  • Tying one’s identity and communications to a single silo means relying on a single point of failure, degrading the overall reliability and stability of the system. (Remember the failwhale and efforts to keep Twitter from going offline during the Iran uprising, for example).

  • Handing over management of one's identity to a company means being dependent on their decisions and priorities. (Consider the 5,000 friend limit on Facebook; Twitter's arbitrary suggested users list; and examples of users being ousted from various services for controversial reasons).

  • A web built on top of a few proprietary platforms means less diversity and ultimately smaller scale than a web built on non-proprietary protocols and standards (consider how useful email, the web, and the internet itself became once open standards for interoperability were adopted, and the power of "small pieces loosely joined").

  • And finally, on an ethical and emotional level — it just doesn't feel right.

Fortunately, there are a number of initiatives that are gaining in popularity and finding pockets of adoption throughout industry, leading us to a juncture, where in one direction is the status quo and in the other is what we call “the people-centric (real-time) web”.

The people-centric (real-time) web

If the document-centric web was dominated by static pages, then the people-centric web is about placing you at the center (as Time Magazine did famously in 2006). We're seeing the rise of dynamic, portable friend lists and non-brand-mediated identities that can be used across a range of standards-compliant websites. People are beginning to move freely between silos. Individuals are increasingly able to bring their data with them and substitute one service or service provider with another, as one can switch between Outlook and Thunderbird for email, or Photoshop and Pixelmator for image editing on the desktop. Relevant information and friends' activities are starting to come to users via distributed push publishing. (Thomas Vander Wal has called this the “come to me” web).

Let us briefly describe the key enablers of this emerging new phase:

Portable profiles means that instead of creating an account on each service you join, you can now host your identity in one place and bring your profile and friends with you to other sites as you surf the social web. Webfinger, OpenID, Portable Contacts, and OAuth all make this possible (and for bootstrapping profiles from the legacy document-web, we have Google's Social Graph API).

Distributed push publishing means there is no longer a need to rely on proprietary platforms. The emerging standards here are PubSubHubbub (PuSH) and rssCloud (see comparisons on TheNextWeb and TechCrunch).

Synchronized conversation threads means that users can participate on the same conversation thread across multiple interfaces and services (we are still waiting for a standard, for which various geeks are actively devising a plan).

Much work remains to make cloud services fully interoperable, but the foundations are in place to turn the web into a truly people-centric place. This call to action goes out to developers, corporations, and individuals alike. Best of all, it's not that hard to start supporting these efforts:

Let people use existing accounts to sign in and sign up for your service. First, the signup ritual offers the least amount of value to users so get it out of the way as fast as possible! Plus, it's an automatic barrier to entry — you'll see an increase in successful signups by reducing the friction in logging in up front (as Plaxo did). Second, unless it's core to what you do, this will also save you the chore of managing profiles on your service. Third, people have so many profiles these days, they can't keep track of them and they certainly don't want to be creating yet another. Instead, figure out a way to subscribe to someone’s existing profile — and keep a reference of it up to date on your site.

Sharing information and activities from your site is how other people will discover you. Stickiness as a business practice was a byproduct of the document era of the web; on the people-centric web, portability is critical. Data, identities, relationships, and activities need to flow between sites in order to expose insights, spread knowledge, and engender meaningful social interactivity. This sounds complicated but is relatively straightforward. To begin, your site can make available atomic units of data, exported as streams of activity that indicate who acted in which way upon what object. It’s easier than it sounds and formats are available to support this modular approach (see: Activity Streams)

As a user, consider how much control and security you really want over your online identity. How do you feel about leasing an identity from a web brand? Unsure about the benefits of owning your own? Some providers (Google, Yahoo, Flickr, MySpace, AOL) let you use their accounts as OpenIDs — a great step towards portability, and beneficial to everyone. The catch with any leased identity is that your identity will be under the provider's brand, profile constrained by their design decisions, and personal data subjected to their terms of service. As an alternative, acquiring your own domain and setting up your own profile with an independent is becoming much easier with free services like and More innovation is needed in this area to make independent identities for people and organizations first class citizens on the social web, and their setup and management simpler, accessible, and secure!

What's yet to come

It’s 2009, going on 2010. For the past three years, the web has been morphing into a real-time and people-centric place. We've seen this trend among individual users — through their actions and demands for better social experiences — but also increasingly among companies and developers. We want a web that's more "like us" than the old model was. We want a web where people are as important to the architecture of the system as documents.

And with this new model come new opportunities for innovation and personalization. It is possible to build applications for participating in decentralized conversations around various ideas and trends. This presents a new opportunity for identity management apps, community sites, social dashboards, real-time search, messaging hubs... and even browser makers, hardware manufacturers, and ad networks. Mobile platforms are also growing, as people connect over non-desktop devices. These small handheld technologies further underscore the importance of portable identity, microcontent, decentralization, and (near) real-time delivery. A document-centric approach just doesn't make sense in a mobile world, and with new ground being broken in fields like augmented reality, demand for increasingly rich social experiences powered by open standards instead of proprietary platforms will continue to grow.

But consider the future: the benefits of a people-centric model are still evolving and remain to be fully realized. It's critical to not be complacent with the platforms we've grown so accustomed to. If you wear the developer's hat, now's the time to get on board, read the specs, and implement support for OpenID, Activity Streams, OAuth, PubSubHubbub/rssCloud, or the other mentioned open standards that are relevant to your users. If you are a user, don't be afraid to be vocal and ask the services you love to show they love you back, by giving you the rights to your data and the tools to take it with you elsewhere. If you're a business, realize that the distributed potential of the social web has barely been tapped, and that you have a choice between (as Robert Scoble calls it) gifting your branding power to someone else, or leveraging these standards to turn your own site from an island to a node in a network of social activity as wide as the web itself. In the end, the internet as a whole will be better off if we stay in control of our own destinies.


Chris and Jyri are presenting a workshop on this material at their MindTrek pre-conference tutorial on September 30th in Helsinki. Early bird tickets are still available at a discounted rate; reserve yours today! Also, don't forget you can still register for MindTrek, the Nordic conference on social media (Oct. 1st–2nd) in Tampere, Finland.

blog comments powered by Disqus
Julien, September 11, 2009

Great article.

Christina Ivazes, September 12, 2009

Great historical evolution of web. I am waiting anxiously for this new technology to catch up to our needs for centralization.Who can keep track of all this? I'd rather be spending my time on productivity rather than trying to remember which sites I need to visit in order to maintain the diverse networking I need to keep business flowing and growing. Thanks! I look forward to more progress updates for non-techies like me.

Deirdre (120new) 's status on Saturday, 12-Sep-09 03:59:15 UTC -, September 12, 2009

[...] a few seconds ago from web [...]

Robert Aarts, September 12, 2009

Nice thought provoking story. I certainly agree with the gist of it.

I can't resist though to point out that a very, very, similar vision of real-time, person/identity centric. privacy aware, constellation of services crystallized into a large set of tried- and tested specifications during the first half of this decade:

Admittedly those specs prescribe somewhat bloaty protocols, but I suggest for everybody interested, or involved, in OpenID, Activity Streams, OAuth, PubSubHubbub/rssCloud, etc., to read the Liberty documents and use them as "requirement specifications". Because it's all there... Only needs to be re-cast into today's fashionable technologies ;)
Just as an example: there is an "template" specification for profiles including protocol to subscribe to (and publish changes in) any (part of any) profile based upon that template specification.

@Christina: the Liberty solution to the (real) problem you describe is the "Discovery Service". This is the web site that knows where your other services are located and can tell other web sites about those (under your control).

Contador Harrison September 12, 2009

The article highlights how far the web technology has come.the dynamism is one thing that tech savvy folks like me have to keep in touch with every hour,week in week out and help the world change the way they live,communicate, socialize,conduct politics and economic activities.

Sorry non tech,the tech is here to grow and stay.there will never be another world without it.

pubsubhubub and rss-cloud : changing the way you read the web « Thoughts on technology and social web, September 12, 2009

[...] Fat Pings Win There’s a Reason RSSCloud Failed to Catch On The Reason RSS Cloud Can Work Now The Web At a New Crossroads Possibly related posts: (automatically generated)Real time roundup – Part 3 : Instant [...]

Konsepti » Arkisto » Henkilökeskeinen verkko, September 14, 2009

[...] Messina ja Jyri Engeström kirjoittivat Arctic Startupin artikkelissaan dokumenttikeskeisen verkon muutoksesta henkilökeskeiseen suuntaan: If the document-centric web was [...]

Our Links from September 13th through September 14th | Socialsquare - we make participation happen, September 14, 2009

[...] Andreas L: The Web At A New Crossroads - Describing the journey from a document-centric to a people-centric web. [...]

Denton Gentry, September 14, 2009

Very nicely written, I thoroughly enjoyed reading it.

One note: also went with a purely OpenID-based account system when it launched in 2008. Its a site for programmers, which is probably why they were more willing to push the envelope a bit before OpenID was really ready.

I like the idea of controlling my OpenID instead of allowing a provider to own the brand. I tried setting up myopenid with my own domain, but haven't managed to find the time to finish it. I'll look at and as well, thanks.

The Web at a New Crossroads | FactoryCity, September 14, 2009

[...] post is a collaborative essay written by Jyri Engström and myself, edited by Brynn Evans and originally posted to the ArticStartup blog on September 11, 2009. Thanks to Brad Fitzpatrick for his comments on the [...]

Jonas delade grejer / 2009-09-15 | You. Me. We., September 15, 2009

[...] The Web At A New Crossroads [...]

Andy Oram (praxagora) 's status on Monday, 14-Sep-09 23:37:39 UTC -, September 15, 2009

[...] [...]

ARJWright, September 15, 2009

As with others who have read/commented here and elsewhere, I totally agree. I have to wonder about my own usage of a mobile web server and this idea of the person being a publishing/connecting point. I need to go one step further, but I see exactly what you are proposing has happening. Very cool.

heimohanninen, September 15, 2009

Good observations. Even me - a bit slow learner of web N++.0 - have changed my reading and information discovery habits the way described here. TweetDeck has become an important "what's happening radar" to me (spotted this too).

Portable profiles I first time used just recently at . A site that addresses the next burning issue – not HOW to share – But WHAT to share and are we talking about the same thing - A subject and further on semantics. Foundation for semantic web is the ability to define, publish and agree at some level at least that what is identity of “a thing”.

We need a capability to easily somehow indicate identity of “a thing” what is the meaning of it (some sort a profiling?) and make it available for other to share, promote or demote etc. … distributed semantic name server platform …. :-/

Scott Morrison, September 15, 2009

Certainly sparks the imagination for things to come. If websites can be crawled and their information freely available so should a persons identity, if they so choose. Problem is having one source, not an infinite, so the metadata can be relied on.

Mike Keen, September 15, 2009

I couldn't agree with you more. Realtime services have been quite common for some time now. The thing that is really going to push the need for high-performance, scalable realtime networks is going to be the interfaces that we've already seen creeping onto the scene. Facebook's friend notifications were the first mainstream sign of this. Keep an eye out for more and more of these types of interfaces, because they're popping up everywhere.

Rebecca Caroe, September 15, 2009

Guys - thanks for a well-thought out piece as ever in a timely fashion. I think there's more to be added to your final paragraph as advice for businesses.

It is more than just "a node in a network of social activity", the shift in power from the organisation to the consumer is what Doc Searls has described as "The Logic is AND not OR" in his foreword to a report on Volunteered Personal Infromation (VPI).

He adds "This shift won’t happen because companies give customers power. Instead this power will grow natively with individuals. VPI is a primary fuel for that growth."

Let me know if you want this material for your MindTrek event.

Sudheer, September 15, 2009

I am excited to travel the web road ahead.

The people-centric web | Mashstream, September 15, 2009

[...] web Tuesday, September 15th, 2009 | Lifestreaming, Real-Time Web | admin According to ArcticStartup, the web is at a new crossroad where it is dominated by “dynamic, portable friend lists and [...]

The Future of Social Media :: Björn Klose, September 15, 2009

[...] Arduino, portable identity (Facebook connect, OpenID), semantic web, Google Wave Applications and [...]

jungleG » Blog Archive » Distributed Social Networking Might Be Dead On Arrival, September 15, 2009

[...] Messina talked about developers on a recent blog post: If you wear the developer’s hat, now’s the time to get on board, read the specs, and implement [...]

Chris Messina, September 15, 2009

Thanks Rebecca. We're big fans of Doc Searls' work with Project VRM, and he and I have had several conversations about these topics.

I'd be very interested to read the report that you've developed — as sourcing evidence for these trends is certainly one necessary aspect of making our case!

Indeed, to reiterate your point, this change is happening because of the way that people are using the network, not necessarily because of any prescribed architecture. That said, of course, certain companies are accelerating these changes based on how they're building out features (i.e. including Facebook and Twitter links into the core UI of their applications), but much of this is emergent.

What's more interesting and relevant — to counter the centralized, big company approach that still seems dominant today — is how the popular companies and brands changes internationally. It's really not kosher to just rely on Facebook and Twitter to be the ONLY networks worth sharing to when you enter into non-US/non-Western markets.

That's where, I think, the decentralized web really shines and demonstrates its necessity.

Chris Messina, September 15, 2009

Apple's push notifications in the iPhone are another — as well as the drive towards the "real-time web" — something that, prior to Twitter, Google hadn't been focused on.

With the rise of ubiquitous connectivity, there's really no reason why information can't be received as fast as it's created — and dealing with interfaces to cope with the deluge of real-time information/noise is an exciting and developing arena...! Will be curious to see what the real-time feed readers of the next 3 years will look like...

Chris Messina, September 15, 2009

Thanks Denton.

Stackoverflow is definitely an interesting case. I think you're right that their audience is more "ready" for OpenID and for the concept of federated identity than the casual web audience.

That said, it is clear that putting any barrier to engagement in front of a user will distract them and cause them to reconsider participating. There is therefore a need in less technical realms to make the "where you're from/who are you" process simpler and more streamlined. In other words, Stackoverflow gets the identity bit out of the way as fast as possible, because its users get "URLs as identifiers". We need more innovation and usability for casual web users, but I believe that momentum is building.

It's also important that technologies like OpenID provide a meme around which people can discuss an idea... prior to OpenID coming on the scene, talking about "federated identity" was hard... but OpenID provides a branded token that lets people compress several concepts into one, in order to talk about what they want and need.

That, in my mind, is critical to seeing wider spread adoption in the open marketplace.

Our links for September 15th | Socialsquare - we make participation happen, September 16, 2009

[...] Martin: The Web At A New Crossroads - The people-centric (real-time) web - If the document-centric web was dominated by static pages, then the people-centric web is about placing you at the center. We’re seeing the rise of dynamic, portable friend lists and non-brand-mediated identities that can be used across a range of standards-compliant websites. [...]

Jonas delade grejer / 2009-09-15 | You. Me. We., September 16, 2009

[...] – The Web At A New Crossroads2009-09-14 – Very [...]

Brian P O'Rourke, September 16, 2009

Great article.

"In the abscess of innovation since the bubble burst" - this is either a typo or a very strange metaphor....

links for 2009-09-15 - NOWUSEIT.COM, September 16, 2009

[...] The Web At A New Crossroads The people-centric (real-time) web – If the document-centric web was dominated by static pages, then the people-centric web is about placing you at the center. We’re seeing the rise of dynamic, portable friend lists and non-brand-mediated identities that can be used across a range of standards-compliant websites. (tags: internet web web2.0 socialnetworking collaboration technology future socialsoftware chrismessina socialsquare) [...]

Forecasting The Future | NavinoT, September 16, 2009

[...] jika Anda ingin membuat next big portal, think again. Social network? Memang people-centric tapi konsep Facebook sepertinya sudah usang dan harus [...]

Ville Vesterinen September 16, 2009


During the editorial process we had 'abscess' in there because of the 'bubble burst' analogy. I think either one is fine, but It might make more sense to say 'absence' if the connection is hard to pick up as a reader otherwise. I made the changes after consulting the authors and Brynn Evans, who edited the piece. Thanks for the feedback.

links for 2009-09-16 « Amy G. Dala, September 16, 2009

[...] The Web At A New Crossroads (tags: technology distributed collaboration web) [...]

090916 What's coming |, September 16, 2009

[...] The Web at a New Crossroads is an informative article that wrestles with the metaphor as architecture problem. It’s an indicator of a better direction. The authors are still trying to break out of the hard constraints imposed by the document metaphor. [...]

newslife site » Navinot: Forecasting The Future, September 17, 2009

[...] jika Anda ingin membuat next big portal, think again. Social network? Memang people-centric tapi konsep Facebook sepertinya sudah usang dan harus [...]

  Bookmarks for August 25th through September 20th by, September 20, 2009

[...] The Web At A New Crossroads – A post that looks at the history of the web and where it may take us. But this is no dry history lesson. This is more a history of the approach to getting information online. A great post [...]

Superfeedr Blog : Time vs. Space : how do we apprehend information? « Blog Looppa, September 21, 2009

[...] story, images are included in the body, etc… I fully agree to that extend with Chris and Jyri’s article about the web being [...]

P2P Foundation » Blog Archive » Beyond siloed proprietary platforms: the new rights of the people-centric web, September 23, 2009

[...] Since 2007, we moved from a document-centric web to a people-centric web, but the new model is not mature yet and limited because of the dominance of siloed proprietary platforms: [...]

Kommenterat av Stenbeck – 2009-09-23 | Next Generation Internet, September 23, 2009

[...] The Web At A New Crossroads [...]

The Bubble of Coolness | SchizoBlog, September 30, 2009

[...] eivät ymmärrä muut kuin päivän seminaariin osallistuneet (eivätkä kaikki hekään). Ihan vain omaksi huviksi ja hyödyksi tein [...]

Chris Messina MindTrekissä identiteetinhallinnasta verkossa –, October 01, 2009

[...] Kannattaa myös lukea Chris Messinan ja Jyri Engeströmin vieraskynä Arctic Startup -blogista: &#82... [...]

Mindtrek’s First Day – Full Of Variety, October 02, 2009

[...] year, especially Chris Messina and Jyri Engeström, who were keynote speakers continuing on the post they wrote for ArcticStartup [...]

The Next great media company won’t have a website « End of Business as Usual – Glenn's External blog, October 17, 2009

[...] This is all very much in the spirit of the Real Time Web and the growing importance of Activty Streams. For more on these topics, see Chris Messina’s posts here and here. [...]

Darren Bounds (darrenbounds) 's status on Saturday, 07-Nov-09 21:36:04 UTC -, November 07, 2009

[...] a few seconds ago from Cliqset [...]

Open the doors, and see all the people. « Infovark, November 09, 2009

[...] Evans pointed me to an intriguing piece of future gazing from Chris Messina and Jyri Engeström – The Web at a New crossroads. In it, they describe the evolution of the internet from a document-centric content sharing [...]

???????? ?? ?????: 8 ??????? – 14 ??????? 2009 : ????? ?????? ????, November 14, 2009

[...] ??? ???????? ????? ?????: ???? ????? ?? ????? ?? ???????? ???? ????? ????, ???? ?????? ?????: The Web At A New Crossroads [...]

Brynn Marie Evans » Why there’s nothing to fear in social search, February 01, 2010

[...] improvement on limited algorithms that only index static pages. Chris Messina and Jyri Engeström remind us that we’re in a transitional state from a Web of documents (in which algorithms were [...]

vesti srbija, March 31, 2010

"It’s 2009, going on 2010. For the past three years, the web has been morphing into a real-time and people-centric place". This is great.

sobre tecnologia » Blog Archive » El valor de la centralización frente al mito de la red social distribuida, May 14, 2010

[...] Algunas lecturas extra para quien quiera seguir con el tema en RWW, Dreig, NIA, Chris Messina y Jyri Engeström. [...]